[sdiy] "zener" noise

[Magnus Danielson]

From: raywilson at comcast.net
Subject: Re: [sdiy] "zener" noise
Date: Mon, 11 Jul 2005 13:38:51 +0000
Message-ID: <071120051338.28782.42D2766B00070C640000706E220588644202019C040799970E9D at comcast.net>

> It can definitely be done. Its just that zeners seem to be a bit more finicky than reversed biased transistors about making a lot of noise. Also I have read that the zener effect is actually different from the avalanche effect. The reversed biased transistor gives an avalanche effect.
> 
> Anyone want to chime in on the difference.

>From the top of my head, the difference is this:

In Zener diode action, the heavilly doped PN-junction is being reversely biased
such that the Fermi energy barriers (top of one layer becoming higher than the
other) such that electrons can tunnel thru the barrier. This is the zener
effect.

In Avalance diode action, the lightly doped PN-junction is becomming
increasingly strained in the biasing such that it eventually experience break-
thru in where one electrons releases two in another atom and thus forming an
avalance chain reaction. This avalance chain reaction probably contribute more
to noise than the tunneling of the zener action, but that is just me guessing.

Anyway, the interesting thing is that there is a cut-off in range of these
effects, such that the zener action is bound to levels below 10V (depends on
doping) where as the avalance action ranges higher voltages. This makes "zener"
diodes below some voltage is actually zener diodes where as for higher voltages
you have avalance diodes. The "Zener" diode branding is a bit confusing in this
context.

When talking about diodes, I just read an article in which they had used a
1N4148 as a x7 harmonic stepup of 100 MHz to 700 MHz with low phase noise
consequence in a cheap GPS antenna setup. Our normal quick-and-dirty diodes is
a better then we usually is beleive them to be. Of course, there is much better
diodes around for that task, but I think they wanted to make a point when doing
a precission tool.

Cheers,
Magnus